Celtic Religion: An Overview

views updated


Historical references to the Celts begin in the fifth century bce. Herodotus and Hecataeus of Miletus are the forerunners of a long series of Greek and Latin writers whose reports and comments, both well- and ill-informed, reflect the changing fortunes of the Celtic peoples during the pre-Christian era and their impact on the Greco-Roman world. Herodotus and Hecataeus confirm that by about 500 bce the Celts were already widely dispersed over central and western Europe, including perhaps Gaul and the Iberian Peninsula, and evidence from the fifth century testifies to further territorial expansion. About 400 bce this process quickened as tribal bands invaded northern Italy and established settlements that, in due course, became the Roman province of Gallia Cisalpina. Some Celtic bands raided farther south, as far as Rome and Apulia and even Sicily, and around 387 they captured and sacked the city of Rome, an event of traumatic importance in Roman history.

To the east, other Celtic tribes penetrated into the Carpathians and the Balkans during the fourth century bce. In 279 some of them entered Greece and plundered the shrine at Delphi, and in the following year three Celtic tribes, known collectively to the Greeks as Galatae, crossed into Asia Minor and eventually settled in the region that still bears the name Galatia. In Britain, the final phase of Celtic settlement came with the arrival of the Belgae in the first century bce, although there is archaeological evidence of earlier immigrations dating back as far as the fifth century bce. For Ireland, the evidence is complicated, and one cannot confidently infer a Celtic presence before the third century bce.

By the early third century bce the Celts extended across the length of Europe from Britain to Asia Minor, and they were considered one of the three or four most important barbarian peoples in the known world. Thereafter, however, their history is one of decline. Harried by Germans in the north, Dacians in the east, and Romans in the south, the continental Celts saw their widespread dominion disintegrate and contract until their realm came to be associated solely with Gaul, where they maintained their independence until their conquest by Caesar (10044 bce) in the midfirst century bce (5851 bce).

In Britain and Ireland the process was longer drawn out, but there too Celtic society was gradually eroded and submerged by foreign domination. By the beginning of the twenty-first century, Celtic languages were being spoken only on the western periphery, in restricted areas of Ireland, Scotland, Wales, and Brittany. The insular languages belong to two distinct branches of Celtic and perhaps reflect an older dialectal division among the Celtic-speaking peoples of Europe: Goidelic, which comprises Irish and Scottish Gaelic (and formerly Manx), and British or Brythonic, comprising Welsh and Breton (and formerly Cornish). However, Breton, which is largely the product of immigration to Brittany from southwest Britain from around the fourth to the seventh century ce, may also have absorbed surviving elements of Gaulish speech.

The entry of the Celts into the written record coincides with the first evidences of the Second Iron Age, also known as La Tène culture, which refers broadly to those areas of Europe historically associated with the Celts. However, the further back beyond the fifth century bce one goes, the more difficult it becomes to use the term Celts with reasonable confidence, because the correlatives of language and written reference are lacking. The cultural phase which preceded La Tène, known as Hallstatt, dates from the ninth century bce and covers an expanse of territory extending at least from Burgundy to Bohemia. Hallstatt culture is characterized by elaborate chariot burials and by the use of iron rather than bronze for arms and utensils. It is the product of a warrior aristocracy that is generally recognized as Celtic, or at least as the direct ancestor of the Celts of the following period. Obviously, the definition of a Celtic identity was the product of a long period of linguistic and cultural evolution, and some archaeologists have ventured to identify as proto-Celtic the peoples of the Urnfield culture and of the Tumulus culture that preceded it in the second millennium bce, or even the peoples of the Beaker and Battle-Axe cultures of the third millennium bce. However, this is mere speculation; the point in the archaeological record at which the Indo-Europeans made their appearance in central and western Europe cannot be known with certainty. And yet most scholars discern in the culture of the Tumulus peoples features that are echoed in that of La Tène.


The sources for Celtic religion fall broadly into two categories. The first category comprises the various monuments relating to the Celts on the continent, particularly in Gaul and in Roman Britain, and the second category comprises the insular Celtic literatures that have been preserved in writing. The two types pose problems that are very different in character. Most dedicatory inscriptions, images of Celtic deities, and commentaries by classical authors belong to the Roman period and probably reflect in varying degrees the effect of Roman influence on Gaulish institutions. For example, because Gaulish sculpture is based for the most part on Greco-Roman models, it is often difficult to assess and interpret its relevance to native belief. Even cases in which motifs and figures seem clearly to derive from pre-Roman religious tradition, as in some of the Celtic coins of the third and second centuries bce, they are not easily related to what is known of insular Celtic myth and ritual.

The difficulty lies in the lack of the literature that would provide a context for the iconography as well as a key to its understanding. The druids, as Caesar records, accorded primacy to the spoken word and refused to commit their teaching to writing. Consequently, the whole of the traditional literature, including the mythology that gave the iconography its meaning, was confined to oral transmission and perished with the extinction of the Gaulish language. The total loss of this vernacular literature, which was doubtless comparable in volume and variety with that of early Ireland, renders all the more significant the testimony of those classical authors who recorded their own or others' observations on the Celts. Probably the most important was Posidonius (c. 135c. 50 bce), who had firsthand knowledge of diverse cultures, including the Celtic in southern Gaul, and who devoted the twenty-third chapter of his lost Histories to Celtic ethnography. Much of his account of the Celts survives in the work of later writers who borrowed from him, such as the historian Diodorus Siculus (died after 21 bce,), the geographer Strabo (c. 63 bce24 ce), and, most notably of all, Julius Caesar, whose account is crucial for the study of Gaulish religion.

The limitations of the classical sources are obvious. Most of the reports come at second- or third-hand and are subject to the prejudices and preconceptions born of classical civilizationor even, as in the case of Caesar, of internal Roman politicsbut they are not without substance, as on many points they harmonize remarkably with the later insular sources. For example, classical sources note that in Gaul there were three classes associated with literature and learning: the druids, the bards, and, between them, an order that seems to have been best known by the Gaulish term *vātis (cognate with Latin vatis ; * denotes a form not appearing in epigraphs and reconstructed from the quotations of Greek and Latin authors), which is not clearly distinguishable from the druids. Far removed in time and space, the same threefold arrangement occurs in medieval Ireland, comprising here druids (druïdh ), filidh, and bards (baird ). The term fáith (prophet) is the Irish cognate of Gaulish *vātis and appears frequently as a near synonym of fili (plural, filidh ).


The second main body of evidence, the insular Celtic literatures, is at first glance far removed from the pre-Roman world of the continental Celts. The great historian of Gaul, Camille Jullian (18591933), questioned whether it was valid to use Irish and Welsh literary sources to interpret Latin and Greek references to Gaulish institutions and concluded that one could not rely on documents written so long after the Celtic migration to Ireland. In fact, the gap is much narrower than the twelve centuries that he supposed, because much of the relevant material is linguistically older than the period of the manuscript collections in which it is now preserved. Further, there is no evidence that Christianity was introduced to any part of Ireland before the second half of the fourth century ce, or that it impinged much on the traditional culture of the country before the sixth century. Moreover, one must reckon with the highly conservative character of Irish learned tradition, which, thanks to the assiduousness of the hereditary filidh, survived far into the Christian period and transmitted innumerable elements of form and content, particularly in the area of social institutions, which find their closest detailed analogues in the sacred texts of Vedic and classical Sanskrit.

Written literature in Irish dates from the second half of the sixth century ce, when monastic scholars adapted the Latin alphabet for that purpose, and it gradually increases in volume during the following centuries. In addition to a good deal of typically monastic learning, both religious and secular, the literature comprises a vast amount of varied material recorded or adapted from oral tradition. However, only fragments of this literature survive in contemporary manuscripts, mostly in the form of annals or notes and glosses accompanying Latin texts; all the vernacular manuscripts written before the end of the eleventh century, some of them known by name, have perished through usage or spoilage caused by warfare. Then around 1100 came Lebhor na hUidhre (The book of the dun cow), probably written in the monastery of Clonmacnois and the first of a series of great vellum manuscript compilations that were part of a conscious endeavor in the face of ominous political and social change to conserve the monuments of native tradition. It was followed around 1130 by an untitled collection now at the Bodleian Library at Oxford University and around 11501200 by Lebhor na Nuachongbála (known commonly as the Book of Leinster), probably compiled in the monasteries of Glendalough and Terryglass, respectively. Over the next couple of centuries a number of major manuscripts appeared, of which the most important are the Great Book of Lecan, Yellow Book of Lecan, Book of Ballymote, Book of Lismore, and Book of Fermoy. These capacious bibliothecae embrace all the various genres of traditional literature: hero and king tales, mythological tales, origin legends, genealogies, onomastic (the study of proper names) and etymological lore, gnomic texts, legal tracts, eulogy and elegy, battle tales, birth tales, death tales, tales of the otherworld, and so on. It is important to remember that, although the surviving manuscripts date from a relatively late period, the matter they contain has generally been copied more or less faithfully from earlier manuscripts. The result is that the initial redaction of the individual texts can be dated with a fair degree of accuracy on the basis of linguistic criteria. Thus the texts are often demonstrably centuries older than the extant manuscripts.

Along with these manuscript collections, several specialized compilations, including Leabhar Gabhála Éireann (The book of the taking of Ireland), commonly known as the Book of Invasions, an amalgam of myth and pseudohistory, which purports to recount the coming of the Gaels to Ireland as well as the several immigrations that preceded it; the Cóir Anmann (Fitness of names), a catalog of names of "historical" personages with many imaginative etymologies and references to traditional legends; and the Dinnshenchas (Lore of famous places), which provides a much fuller and more elaborate examination of place names than the Cóir Anmann provides for personal names. The features of the Irish landscape and their names, if properly construed, were thought to reveal the history of the country and its peoples from their beginnings. From the first shaping and definition of the landthe clearing of plains, the creation of rivers and lakes, and the assigning of names (as related in Leabhar Gabhála )each place was linked indissolubly to momentous events by an association that conferred on it an enduring psychic resonance. The onomastic element is pervasive in Irish (and Welsh) literature, and in poetic tracts dating from around the tenth century, the history of dinnshenchas is included in the course of study prescribed for apprentice filidh. During the eleventh and twelfth centuries, a period of intensive compilation, a comprehensive volume of these onomastic legends was assembled. This mythological gazetteer of Irish place names exists in several recensions (critically revised texts that use varying sources), both prose and verse. Among the many other miscellaneous sources are the lives of the saints, particularly those later ones compiled or redacted from the eleventh century onward (of which it is sometimes said that they contain more pagan mythology than Christianity).

Evidence indicates that the early oral literature of Wales was comparable in volume and variety with that of Ireland. Unfortunately, because of a weaker scribal tradition, Welsh literature is less well documented for the pre-Norman period, prior to the eleventh century. This applies particularly to prose, which in the Celtic languages is the standard medium for narrative and hence for most heroic and mythological literature. Of the compositions ascribed to the fathers of Welsh poetry, Taliesin and Aneirin, who belonged to the second half of the sixth century, only a modest proportion is likely to be authentic, and all of that consists of eulogy and heroic elegy. However, from the ninth or tenth century onward Taliesin became the focus of poems and stories (extant only in much later versions) that represent him as a wonder child, seer, and prophet; some of these motifs clearly derive from native mythological tradition. There is no evidence of written Welsh narrative prose before the eleventh century, the period to which most scholars assign the first redaction of the earliest of the group of tales known as the Mabinogi or Mabinogion. However, the earliest manuscripts containing this prose material date from considerably later. Apart from two manuscript fragments from the late thirteenth and early fourteenth centuries, the main texts are the "White Book of Rhydderch" from the midfourteenth century and the "Red Book of Hergest" from the late fourteenth or early fifteenth century. Another important source is the Trioedd Ynys Prydein (The triads of the island of Britain), which contains numerous references to mythological as well as historical characters and events; it may have been compiled in the twelfth century, but much of the contents must have existed in oral tradition before then. Also of mythological interest are the poems compiled as part of the "Black Book of Carmarthen" in the mid-thirteenth century, some of the contents of which may be dated on linguistic grounds to the ninth or tenth century.

Given the diversity of these sources, it is unrealistic to expect from them a clear image of religious and mythological unity. On one hand, Gaulish epigraphy and iconography belong preponderantly to the period of Roman domination when native religion was being progressively modified by Roman influence. On the other hand, the insular literatures, although exceedingly conservative in many respects, were recorded and redacted by monastic scribes and scholars who, however well disposed toward their own vernacular tradition, were nonetheless educated Christians, who on matters of crucial importance doubtless gave priority to Christian teaching over pagan tradition. In short, the integral tradition as it would have been transmitted and commented on by the druids in an independent Celtic society does not exist. Even among the insular Celts, history created important disparities. For instance, Ireland escaped the immediate physical presence of Rome, which left its imprint so clearly on medieval language and thought in Britain and Wales. One must also acknowledge the imponderable but obviously considerable survival of pre-Celtic religion in Celtic belief and practice in the several areas of Celtic settlement. Yet, despite these sources of dissimilation, the underlying structural and thematic unity of British and Irish ideology is more striking than the superficial differences.


The plastic art of the Celto-Roman period is so evidently based on that of Rome that it might appear at first glance to have been borrowed whole and unchanged, but on closer scrutiny it reveals many elements that derive from the Celtic rather than from the Roman tradition. On one hand, there are forms quite foreign to classical art, such as the tricephalic (three-headed) god, the god with stag's antlers, and the god depicted in the Buddha-like cross-legged position. On the other hand, there are images more or less in the classical mode but with features not associated with the corresponding deities of Greco-Latin religion: the wheel, for instance, or the mallet. The wheel is seen by some as representing the thunderbolt, by others as representing the sun, and in some cases it may also be the emblem of the god of the underworld. Similarly, the mallet or hammer is thought to have several connotations: it symbolizes thunder and the sky from which it emanates, but it also functions as an apotropaic (able to prevent evil or bad luck) symbol and as the emblem of an underworld god of fecundity. The cornucopia, or horn of abundance, is not particularly Celtic, but it appears as a common attribute of the Celtic mother goddess, perhaps the most important divinity of the primitive Celtic pantheon. Animal horns are commonly regarded as signs of fertility, and the antlers that the Celtic deity wears on the Gundestrup Caldron, a first-century bce vessel found in Denmark, and elsewhere are taken to symbolize his power and fecundity. Another frequent emblem of divinity is the ornamented torque, which is interpreted to denote a powerful god who is able to provide protection from evil spirits. Although it is usually worn around the neck as a metal collar, the torque is sometimes held in the hand, and, on the relief of the Celtic god Cernunnos in the Musée de Cluny in Paris, the deity carries two torques suspended on his horns.

Probably the most notable element in the religious symbolism of the Celts is the number three; the mystic significance of the concept of threeness is attested in most parts of the world, but it seems to have had a particularly strong significance for the Celts. This is confirmed both by Celto-Roman iconography, which has its three-headed and three-faced deities (and even a triphallic Mercury) and its triads of mother goddesses, and by the insular literary tradition, which has an endless variety of ternary groups in which the triad is an expressive restatement of an underlying unity. Examples include goddesses such as the three Brighids and inseparable brothers such as the three companions of the tragic heroine Deirdre. It is commonly accepted that ternary repetition has an intensifying force, expressing totality or omnipotence, although its symbolism may be even more complex and subtle.

Continental Deities and Insular Equivalents

Given that the bulk of the relevant evidence belongs to the Roman period, the Gaulish religion is for the most part as seen through Roman eyes, which means that it is perceived and presented in terms of Roman religion. A classical example is the passage in Caesar's Gallic Wars in which he lists and defines the principal gods of the Gauls:

Of the gods they worship Mercury most of all. He has the greatest number of images; they hold that he is the inventor of all the arts and a guide on the roads and on journeys, and they believe him the most influential for money-making and commerce. After him they honor Apollo, Mars, Jupiter, and Minerva. Of these deities they have almost the same idea as other peoples: Apollo drives away diseases, Minerva teaches the first principles of the arts and crafts, Jupiter rules the heavens, and Mars controls the issue of war. (Gallic Wars, 6.17)

What Caesar offers us here is a thumbnail sketch of the Gaulish pantheon modeled on that of Rome. As part of this glaringly Roman interpretation, he refers to each deity not by his proper Celtic name but by that of a Roman deity to which it is most easily equated. At the same time he introduces a neat schematism, which is quite foreign to all that is otherwise known of Celtic religion. In thus equating gods and divine functions that are not really equal, he has posed many problems for modern scholars who seek to identify Caesar's Roman gods in continental Celtic iconography and insular Celtic mythology.

To confound matters further, modern scholars have tended to depreciate Caesar's testimony on the Gauls; first, on the grounds that he distorted the facts to enhance his own achievements, and second, on the grounds that he took his information from Posidonius, but used it inaccurately. It has been argued, for example, that Caesarand even Posidoniusexaggerated the social and political importance of the druids, assigning them a dominant role that they never in fact possessed. Yet in this regard, as in others, Caesar's version of things is largely confirmed by the independent evidence of the insular literatures. Once allowance is made for the synoptic nature of his comment, his inevitable professional bias, and the limitations of his interest in Gaul, there is no reason to assume that his account is not largely authentic. By the time he wrote his account, he had had eight years' experience of the country, and most likely he derived much of his information from personal observation and from the reports of colleagues and acquaintances; certainly there is little basis for the common assumption that he was totally indebted to Posidonius for his knowledge of the land and its people.

The concise precision of Caesar's testimony makes it difficult to correlate with other evidence. Georges Dumézil (18981986) remarked that one of the many traits the early Irish shared with the Indians is that they were both fond of classification and careless of order. The result is that Irish literature is often a curious mixture of meticulous detail and incoherence that finds its closest parallel in some of the Indian epics. One must therefore adjust one's mental perspective considerably as one moves from Caesar to the vernacular literatures. It may be that something of this prodigal disorder is reflected in the continental Celtic iconography, which may help to explain why identifications with Caesar's deities are often more a matter of speculation than of demonstration. But perhaps a more important consideration is that Caesar's account and the iconography refer to quite different stages in the history of Gaulish religion. Periods of profound cultural and political change often bring into prominence popular forms of belief and practice that have hitherto been concealed by the dominant orthodoxy. It seems probable that the religion represented in Gallo-Roman plastic art was less clearly structured and delimited than that maintained by the druids in the days of independence before Caesar's conquest.

Modern scholars have often noted, and sometimes exaggerated, a discrepancy between Caesar's account and the Gallo-Roman evidence, claiming that the evidence does not substantiate Caesar's account of a pantheon of major deities who were worshiped throughout Gaul. In Gallo-Roman dedications, deities may be assigned a Roman name, a native Gaulish name, or a Roman name accompanied by a native epithet. The last two cases clearly have to do with indigenous gods, and even the first group may also. For example, the numerous statues and reliefs of Mercury in the guise of the Greco-Roman god might have been intended to honor that god, but equally they might have been intended to honor a native god by borrowing the classical form together with the classical name. Indeed, many of these images have certain features that betray their essential non-Roman character. It has been observed that the great majority of the several hundred names containing a Gaulish element occur only once. Those that occur more frequently tend to do so in regional or tribal groupings, and many of them have a clear local reference (e.g., Mars Vesontius pointing to Vesontio and Dea Tricoria referring to a goddess of the Tricorii). The inference drawn by some scholars, including Joseph Vendryes and Marie-Louise Sjoestedt, is that, although the Celts had a multiplicity of gods, their cults were local and tribal rather than national. Scholars also cite Lucan's (3965 ce) mention of the deity name Teutates, which they interpret as "God of the Tribe" based on the etymologies of Celtic word *teutā (tribe) and an oath formula from Irish hero tales, Tongu do dia toinges mo thuath (I swear to the god to whom my tribe swears).

But this evidence is susceptible of a different interpretation. A large proportion of the Gaulish forms attested in dedications are mere epithets or bynames; even of those that may be taken to be proper names, it would be quite erroneous to suppose that each indicates a separate deity. As Dumézil remarked in Dieux des Indo-Européens (1952), the names of deities are easily reinvented, and the insular literatures offer examples of major gods known by several different names. As for the form Teutates, it may be a title linking the god to the tribe but does not necessarily confine him to it. By the same token, in early Irish law the small tribal kingdom, the tuath (from *teutā ), was the unit of jurisdiction, and rules of law were explicitly stated to apply i tuaith (within a tuath ). Presumably, then, laws originally applied with equal validity only between members of the same tribe; however, substantially the same lawformulated by the same learned class of jurists related to the druids and filidh was common to all the tribal kingdoms. Similarly, in primitive Ireland the vital ritual of inauguration was founded in the first place on the small tribal kingdom (tuath ), as is enunciated in the law tracts, but it is also replicated at different levels throughout the wider cultural community. And as for the alleged lack of great divinities common to all the Celtic peoples, this is gainsaid even in terms of nomenclature by such insular gods as Lugh and Brighid and their continental equivalents. In short, there is a growing awareness that, despite its all too obvious complexities, the seeming throng of Celtic gods is both less amorphous and more universal than was formerly believed.

Another criticism levelled at Caesar is that he assigned separate functions to the several Gaulish deities in contradiction of the evidence. Some scholars hold that the deities were polyvalent (they can be understood in more than one way) tribal gods, and that to seek to restrict them to distinct spheres of activity is pointless. Others hold that all the various attested gods may be reduced ultimately to a single deity who is both polyvalent and polymorphic (i.e., taking more than one form). Thomas F. O'Rahilly, one of the two principal exponents of this view, believed that the core of Irish and Celtic mythology was the conflict in which this universal deity was slain by a youthful hero using the god's own sacred weapon, the thunderbolt. Pierre Lambrechts, the other principal exponent of this view, believed that originally Celtic religion was bound up with one great deity, possibly a ternary (three-formed) deity endowed with multiple and comprehensive attributes and that during the Roman period this largely undefined and impersonal deity was fragmented into a number of smaller, specialized deities through contact with the Greco-Roman world.

This notion of a single all-encompassing god, endlessly varied in form and function, has perhaps a certain plausibility. Because the Celtic gods were not clearly departmentalized, it is difficult to pair them off neatly with their Roman counterparts, and so one finds such evident anomalies as the occasional use of the same Gaulish byname (e.g., Iovantucarus and Vellaunus) with different Roman deity names (e.g., Mars and Mercurius). However, although the functional roles of the several deities are not clearly defined and delimited and frequently overlap with one another, it does not follow that they may be reduced to a single, all-purpose divine overlord. It has often been remarked that in polytheistic systems each god tends to move beyond his or her normal functional field toward a kind of universalism. Yet, despite this tendency toward the assimilation of roles, the insular Celtic gods are far removed from functional indifferentism, and there are some, like Goibhniu (The Smith) and Dian Cecht (The Leech) whose central responsibilities are defined very precisely. The assumption of undifferentiated polyvalence that underlies the conflicting interpretations of Vendryes and O'Rahilly (i.e., tribal and polytheistic) or Lambrechts (i.e., vaguely monotheistic) has not been substantiated. In fact, more recent scholars, notably Françoise Le Roux and Anne Ross, have moved in the direction of a typological classification of the gods based on criteria of function. The scheme put forward by Le Roux is in close conformity with the principles established in Dumézil's functional theory of Indo-European mythology. Indeed, it could be argued that this typological approach had already been anticipated by Caesar in his brief account of the characteristic activities of the major Gaulish deities.

Mercury or Lugh

Caesar's observation that Mercury was the deity with the greatest number of images in Gaul is confirmed by the surviving evidence of inscriptions, stone statues and reliefs, bronze statuettes, and terra-cotta figures. His image often appears in the mode of the classical Mercury: youthful, naked, and beardless; equipped with caduceus (rod entwined with a pair of snakes), petasos (wide-brimmed hat), and purse; and accompanied by cock, ram, or tortoise. But his image is also found in Gallo-Roman guise: mature, bearded, and dressed in a heavy cloak. Sometimes, as in the east and the north of Gaul, he has three heads. Unlike his Roman counterpart, he has a frequent consort named Maia or Rosmerta (The Provider) and includes the art of war in his range of competence.

One cannot assume that Caesar's Mercury coincides with a single native deity throughout the Celtic areas, but there is quite strong evidence for identifying him substantially with the Irish god Lugh (although some doubts have been expressed in this regard by Bernhard Maier). First, Lugh's name and cult were pan-Celtic. Further, Caesar speaks of Mercury as omnium inventorem artium (inventor of all the arts), a close paraphrase of Lugh's sobriquet in Irish, (sam)ildánach (skilled in many arts together). In fact, an episode in the tale of the mythological Battle of Magh Tuiredh dramatically sets forth Lugh's claim as the only god who was master of all the arts and crafts. At Osma in Spain an inscription was found with a dedication on behalf of a guild of shoemakers to the Lugoves, whose name is the plural of Lugus, an older form of Lugh. Most likely these divinities, who recur in an inscription from Avenches in Switzerland, are simply the pan-Celtic Lugus in plural, perhaps triple, form. The Middle Welsh tale Math vab Mathonwy may well echo this connection with shoemaking, for Lleu, the Welsh cognate of Lugh, operates briefly as a high-class practitioner of the craft.

In Ireland, Lugh was the youthful victor over malevolent demonic figures, and his great achievement was to kill the cyclopean Balar with a slingshot. Lughnasadh, his feast, was a harvest festival, and at least two of its principal sites, Carmun and Tailtiu, were the burial places of goddesses by the same names, who were associated with the fertility of the earth (as was, apparently, the Gaulish Mercury's consort Rosmerta). Lugh was the divine exemplar of sacred kingship, and in the tale Baile in Scáil (The Phantom's Vision) he appears seated in state as king of the otherworld and attended by a woman identified as the sovereignty of Ireland, reminiscent of Rosmerta. His usual epithet, lámhfhada (of the long arm), relates to his divine kingship. In the Christian period Lugh survived in the guise of several saints known by variants of his nameLughaidh, Molua, and othersand the motif of the arm is reflected in these Christian traditions as well.

Gaulish Mars

A famous passage in Lucan's (3965 ce) Civil War refers to the bloody sacrifices offered the three Celtic gods: Teutates, Esus, and Taranis. A later commentator on Lucan clearly illustrates the difficulty of identifying individual Gaulish and Roman gods, for one of his two main sources equated Teutates with Mercury, the other with Mars. But if, as seems likely, teutates is primarily a title ("god of the tribe") rather than a name, then such confusion is explainable: the god of sovereignty and the arts, Mercurius, will also function as a warrior, whereas the god of war, Mars, will often function as the protector of the tribe. Consequently, their functions will sometimes overlap, and it may be a matter of chance or circumstance which is given preeminence in a given time or place. A further complication is that many of the Gallo-Roman dedications to Mars present him not only as a god of war but also as god of healing and guardian of the fields, but this may reflect an extension of his role in the Roman period and does not necessarily discredit Caesar's description of him as god of war. So far as the insular tradition is concerned, a god of war does not come into clear focus, perhaps because fighting is a more or less universal rather than a differentiating feature in the heroic context. Thus one cannot easily define the role of Mars, and one cannot so easily assign him a pan-Celtic identity as one can Lugh.

Gaulish Apollo

The classical form of Apollo in Romano-Celtic monuments only partly conceals the several native deities who have been assimilated to him. The use of the plural is probably justifiable, because several of the fifteen or more epithets attached to Apollo's name have a wide distribution, which might suggest that they were independent gods. Yet some of these epithets may have referred to a single deity. Belenus was especially honored in the old Celtic kingdom of Noricum in the eastern Alps, as well as in northern Italy, southern Gaul, and Britain. The solar connotations of the stem bel- (shining, brilliant) would have confirmed the identification with the Greco-Roman Apollo. Grannus, whose name is of uncertain etymology, has a widespread cult with one of its principal centers at Aachen. He is sometimes accompanied by a goddess named Sirona. Borvo, or Bormo, whose name denotes boiling or seething water, is associated with thermal springs, as at Bourbonne-les-Bains and other sites named after him. His consort is Damona (Divine Cow) or Bormana.

This association of healing with springs and wells, which was subsequently taken over into Christian or sub-Christian usage throughout the Celtic countries, tended to encourage localized cults, and it is all the more remarkable that these early names had such an extensive currency. Unlike those already mentioned, Maponos (Divine Son/Youth) occurs mainly in northern Britain, although it is also attested in Gaul near healing springs. Maponos appears in medieval Welsh literature as Mabon, son of Modron, that is, of Matrona (Divine Mother), eponymous goddess of the river Marne in France. A brief but significant episode in the tale of Culhwch and Olwen casts him in the role of hunter and alludes to a myth attested elsewhere in insular literature of the youthful god carried off from his mother when three nights old. That his legend was once more extensive in oral tradition than appears from the extant literature is borne out by the survival of his name into Arthurian romance under the forms Mabon, Mabuz, and Mabonagrain.

His Irish equivalent was Mac ind Óg (Young Lad/Son), otherwise known as Oenghus, who was believed to dwell in Bruigh na Bóinne, the great Neolithic and therefore pre-Celtic, passage grave of Newgrange. He was the son of Daghdha, chief god of the Irish, and of Boann, eponym of the sacred river of Irish tradition (Boyne, in English). As his name and relationship suggest, he is a youthful god, and, perhaps in keeping with this, he is often treated with a certain affection in the literature, particularly in his familiar roles of trickster and lover. But he is nowhere presented as a god of healing, which merely underlines the impossibility of exactly equating Celtic and Roman gods in terms of their functional range.

Gaulish Minerva: Irish Brighid

The goddesses of insular Celtic tradition are involved in a wide range of activities that are only partly reflected in Caesar's succinct comment that Minerva concerned herself with teaching "the first principles of the arts and crafts" (Minervam operum atque artificiorum initia tradere ), even though expertise in arts and crafts enjoyed high status in Celtic society and covered a broad swathe of competences. It is very probable that Caesar chose a single widely revered deity to represent the whole category of goddesses, national and regional. Dedications to Minerva are found throughout the Celtic areas of the continent and in Britain. At Bath she was identified with the goddess Sulis who was worshiped there in connection with the thermal springs and has been identified as a solar deity. The name Minerva is frequently accompanied by the epithet belisama (very brilliant), which suggests a rapport with the Gallo-Roman Apollo, who is sometimes named Belenus (The Shining One). The related plural suleviae is applied to triads of mother-goddesses at sites on the Continent and in Britain. Sulis Minerva is also related to the widespread and important category of mother-goddesses: Matres Suleviae and Suleviae Iunones.

In the Irish context the single goddess who answers best to Caesar's Minerva by virtue of her functional repertoire and wide-ranging cult is the goddess Brighid (from earlier *Brigentī). According to the Glossary of Cormac mac Cuilennáin (c. 900) she was the daughter of the father-god, the Daghdha (literally, Good God), and was worshiped by the filid, the exclusive fraternity of learned seer-poets. In keeping with the Celtic penchant for triadic repetition, she had two sisters, also called Brighidthe one associated with healing, the other with the smith's craftand their combined fame was such that among all the Irish a goddess used to be called Brighid (a statement that invites comparison with Caesar's use of Minerva as an inclusive term for the goddesses of Gaul). Thus, Brighid was patroness of the artistic inspiration of the poets as well as of healing and craftsmanship. Minerva, for her part, is associated with healing, as at the shrine of Bath, and she is also combined on reliefs with Mercury, the master of all the arts, and Vulcan, more specifically connected with the craftsmanship of the smith. It seems clear that Brighid is merely the Irish reflex of a pan-Celtic deity. Her name, which meant originally "The Exalted One," has its close linguistic correspondent in *Brigantî, latinized as Brigantia, the name of the tutelary goddess of the Brigantes, who formed an important federation in northern Britain. She has also a remarkable Christian (or Christianized) double in the person of her namesake Brighid, the great sixth-century abbess of the monastery of Kildare. The legend of the saint is inextricably fused with that of her pagan alter ego, and as she is inevitably accorded a much fuller documentation by monastic redactors, there is the curious irony that the richest source for the mythology of the goddess is the hagiography of the saint together with the prolific folklore that commemorates her in popular tradition. Both the saint's Lives and her folklore suggest a close connection with livestock and the produce of the soil, and, appropriately, her feastday, February 1, coincides with Imbolg, the pagan festival of spring. In a passage of the Topographia Hiberniae that evidently draws on this conflate tradition, the twelfth-century Norman cleric Gerald of Wales (c.1146c.1223; also known Giraldus Cambrensis) reports that Brighid and nineteen of her nuns at Kildare took turns in maintaining a perpetual fire surrounded by a hedge within which no male might enter. Also, it is a significant coincidence that already in the third century Iulius Solinus, associating Minerva with the healing springs of Sulis, mentions in Collectanea Rerum Memorabilium that perpetual fires burned in her sancuary also. In secular texts Brighid is sometimes made to aid and encourage the men of Leinster when they were engaged in crucial conflicts, a reflection perhaps of her pristine role as territorial goddess like those other Celtic deities indicated by such nicknames as Dea Tricoria of the Tricorii in the Narbonnaise, Dea Nemetona of the Nemetes in the Rhine region, or even Dea Brigantia of the British federation.

Celtic Vulcan

Although Caesar does not mention a Gaulish Vulcan, his cult was evidently known to all the Celtic peoples; indeed, the evidence suggests that he enjoyed a higher status than his Roman counterpart. Because he functioned as a very specialized deity, there is a strong probability that his native name among the continental Celts made reference to his craft, as it did in Ireland and Wales, where he was known as Goibhniu and Gofannon, both names derived from the word for smith. The weapons Goibhniu forged with his fellow craft gods, Luchta the Wright and Creidhne the Metalworker, were unerring in aim and fatal in their effect. Further, those who attended the Feast of Goibhniu and partook of the god's sacred drink were thereby rendered immune to age and decay. He was known for his healing powers, and he is invoked in an Old Irish charm for the removal of a thorn. Until the nineteenth century, and in some areas even into the twentieth century, the country smith was still believed to retain something of his ancient preternatural faculty, and he was constantly called on for the healing effects of his charms and spells. In the early tradition, Gobbán Saer (Gobbán the Wright; Gobbán is a hypocoristic form of Goibhniu) was renowned as a wondrous builder, and under the modern form, Gobán Saor, he is the skillful and resourceful mason who outwits his rivals and enemies by his clever stratagems.

Gaulish Hercules or Irish Oghma

Hercules is well represented in Celto-Roman iconography and has a number of regional epithets assigned to him. Doubtless his popularity derives largely from his identification with native Celtic gods who correspond approximately to his classical character. One of these is mentioned in a curious passage by the Greek writer Lucian in the second century ce, who, when describing a Gaulish picture of Hercules, notes that the Celts call him Ogmios. It showed him armed with his familiar club and bow but pictured him uncharacteristically as an old man, bald and gray with his skin darkened and wrinkled by the sun. He pulled behind him a willing band of men attached by slender chains that linked their ears to the tip of his tongue. The explanation, according to Lucian's Gaulish informant, was that eloquence reaches its apogee in old age: the Celts did not identify eloquence with Hermes, as did the Greeks, but with Hercules, because he was by far the stronger.

A question much debated is whether this hoary champion can be identified with the Irish god Oghma, despite the fact that the phonological correspondence is not exact. The functional parallel is adequate: Not merely is Oghma known as a trénfher (strong man, champion), but he is also credited with the invention of the Ogham letters. This system of writing was based on the Latin alphabet and can hardly be older than the fourth century ce, but it probably replaced an older system of magical symbols of the same name.

Gaulish Dis Pater or Irish Donn

Caesar mentions Dis Pater separately from the other gods and states that all the Gauls believed with their druids that they were descended from him. The reference is brief but is sufficient to indicate at least an analogy between the Gaulish god of the dead and his Irish counterpart Donn (Brown/Dark One), whose dwelling place was a small rocky island off the southwest coast of Ireland known as Tech nDuinn (House of Donn). Its English name, the Bull, echoes its other name in early Irish, Inis Tarbhnai (Island of Tarbnae). Tarbhnae derives from tarbh (bull), which perhaps suggests a connection between the god Donn and the great brown bull (the Donn) of Cuailnge, which provides the central motivation for the saga Táin Bó Cuailnge (The cattle raid of Cuailnge).

In his role as god of death, Donn is a rather retiring figure in the early literature. Like Dis Pater, he seems to stand apart from the other deities, but his importance is confirmed by his status in modern folk tradition, in which he is represented as the underworld god who creates storms and shipwrecks but also protects cattle and crops. Both early and late sources record the belief that the dead made their way or were ferried to his island after death. As one early text makes clear, these travelers were regarded as Donn's descendants returning to their divine ancestor. The parallel with Dis Pater is evident and is a further argument for the general authenticity of Caesar's account of the Gaulish deities. Donn's importance in indigenous religious tradition is implicitly recognized in the fact that he is included in the pseudo-history of Leabhar Gabhála Éireann as chief of the Gaels, the Sons of Míl, last of the several peoples to settle in Ireland, but his religious significance presented a problem of how to accommodate him within what was essentially a project of Christianizing native mythic history. The solution the redactors opted for was to dispose of him by having him drown in the sea off the southwest coast and be subsequently brought for burial to a rocky islet nearby that has been known ever since as the Island of Donn.

Sucellus and Nantosvelta

Some two hundred monuments, mostly in Gaul, show a deity holding a hammer, and a number name him as Sucellus (The Good Striker). Besides the characteristic hammer or mallet, he is often depicted with a cask or drinking jar and accompanied by a dog. He is sometimes paired with the goddess, Nantosvelta, whose name suggests an association with water (cf. Welsh nant, meaning brook ). Particularly in the Narbonnaise, Sucellus is frequently assimilated to the Roman Silvanus, guardian of forests and patron of agriculture. Because of these associations and attributes, he has been seen as controlling fecundity, not an unusual function for an underworld deity. He has also been equated with the Celtic Cernunnos and the Irish Daghdha, but although there are certain broad similarities between them, the evidence does not suffice to prove a closer con-nection.

Goddesses and divine consorts

In continental iconography, the frequent pairing of god and consort represents the goddesses as complementary to the male deities, and this image may overlap with the ideal coupling of king and territorial goddess so widely portrayed in medieval Irish literature.

It seems impossible to draw any clear distinction between specific named goddesses and the matres or matronae who appear so frequently in Celtic iconography, often in triadic form like the goddesses of Irish tradition. Both goddesses and matres are concerned with fertility and with the seasonal cycle of the earth, and the insular goddesses are sometimes identified with the land and cast in the role of its protective deities. This intimate connection with the land and its physical features is reflected in the exceptional importance of the feminine element in the dinnshenchas, the vast accumulation of prose and verse, which constitutes a virtual mythological topography of Ireland. A goddess's concern for the land in general also becomes a responsibility for the particular region or kingdom with which she is especially associated. Each goddess ensures the material well-being, sovereignty, and physical security of her particular domain, just as Brighid, in the guise of her saintly namesake, protects Leinster both as goddess of war and as goddess of peace. The mother-god specifically titled as such, Mâtrona, gave her name to the river that is now the Marne in France. She was the mother of Maponos (The Youthful/Son God) known in Welsh as Mabon, son of Modron. In Irish tradition the corresponding role belonged to Boann, eponym of the river Boann (anglicized Boyne); she was the mother of the Irish divine youth par excellence, Mac ind Óc, whose name is the semantic equivalent of the Welsh and Celtic Mabon/Maponos. As mother, the goddess is sometimes represented in Irish texts as ancestress of a distinguished line of descent, and this is presumably what is intended by the author of the medieval Welsh tale "Branwen Daughter of Llŷr" in which he describes Branwen as one of the three great ancestresses of the island of Britain.

In keeping with their titleMatres, Matrae, Matronaethe mother-goddesses attested throughout the Romano-Celtic world are characteristically represented with the various symbols of their maternal and creative function: carrying or caring for infants or bearing such familiar symbols of prosperity as the cornucopia or the basket of fruits. They were also thought of as nourishing and watching over specific peoples and regions and were named accordingly the Matres Glanicae at Glanum (Saint-Rémy-de-Provence), for example, or the Matres Treverae among the Treveri. They would seem to have survived cultural and religious change in the guise of the mamau (mothers) and the formidable cailleacha (old women) of Welsh and Irish-Scottish popular tradition respectively.

Nature associations

Underlying the tradition of dinnshenchas is the belief that prominent places and geological features throughout Ireland were the scene of mythic events or the abode, even the embodiment, of mythic personages. Many of the numerous women who populate this world of onomastic legend are clear reflexes of the multifaceted goddess whose origins are bound up with the physical landscapefigures like Tailtiu and Carmun whose burial places were named after themwere the sites of great royal assemblies. In most of the Celto-Roman world the early onomastic lore disappeared with the indigenous languages, but something of it remained in the divine nomenclature of these areas.

Apart from the general cult of the earth goddess, an extensive repertory of deity names attached to individual places or topographical features also exists. Hilltops and mountain tops are considered particularly appropriate settings for the sacred, as evidenced by dedications to Garra and Baeserta in the Pyrenees and to Vosegus in the Vosges. There was a god of the clearing or cultivated field (Ialonus), of the rock (Alisanos), of the confluence (Condatis), of the ford (Ritena), and of the fortified place (Dunatis). Water, particularly the moving water of rivers and springs, had its special deities, which were generally female in the case of the rivers. One can perhaps glimpse the lost mythology of such rivers as the Seine (Sequana), the Marne (Matrona), and the Saone (Souconna) through the legends of insular equivalents like the Boyne (Boann). The names of many rivers throughout the Celtic lands, such as the French Dives or the Welsh Dyfrdwy, are derived from the stem dev- and mean simply "the divine one." Sacred springs are deified as, for example, Aventia (Avenches), Vesunna (Périgeux), and Divona (Cahors). Further, there were many divine patrons of thermal waters, such as the god Borvo, and this particularly widespread cult is reflected in the countless holy and healing wells (some twelve hundred in Wales alone, and no one has yet added up the Irish instances) that made the transition from paganism to Christianity with little essential change. However, the abundant material evidence for this pan-Celtic phenomenon is not matched by the early insular literary evidence: many Irish tales mention wells with preternatural powers and associations, but there is hardly anything about healing wells as such. Unless this is due to suppression by the monastic redactors of the literature, the only explanation would seem to be that the frequenting of healing wells had always been regarded, even in pagan times, as a popular practice to be distinguished from the more official tribal cults, or simply that it was so familiar as to be unremarkable.

In many instances the holy wells of the Christian period stand close to a specific tree that shares their supernatural aura. Obviously, this is one aspect of the widespread cult of sacred trees. In the Pyrenees there are dedications to the beech (Deo Fago) and to the Six Trees (Sexarbori deo, Sexarboribus) and at Angoulême to the oak (Deo Robori). The Romano-Celtic name of the town of Embron, Eburodunum, contains the name of the deified yew tree. Such continental forms are supplemented by a vast dossier of insular evidence. There were, for example, scores of Christian foundations in Ireland evidently located on the sites of pagan cult centers, each with its sacred tree nearby. The literature frequently mentions several great trees that were particularly honored in tradition: the Tree of Tortu (an ash), the Oak of Mughna, the Yew of Ross, the Bough of Dathí (an ash), the Ash of Uisnech, among others. There was even a special term for such trees, bile, and this term was sometimes used for the great tree that marked each of the inauguration sites of tribal and provincial kings. Standing theoretically at the center of its kingdom like the axis mundi in its greater cosmos, the bile symbolized the integrity and independence of the kingdom. When it happened, as it did occasionally, that it was attacked and felled by a hostile neighbor, this doubtless dealt a severe blow to communal pride and self-respect.

Zoomorphic gods

Celto-Roman iconography contains a rich abundance of animal imagery, frequently presenting the deities in combinations of zoomorphic and anthropomorphic forms. Already noted is the probable connection between Donn, the Irish Dis Pater, and the bull of the same name in the epic Táin Bó Cuailnge. Neither of the two bulls whose conflict forms the climax of the tale is of natural origin. According to other texts, they had previously undergone many metamorphosesas ravens, stags, champions, water beasts, demons, and water wormsand in the beginning they had been the swineherds of the lords of the otherworld. This kind of shape shifting, a continuing expression of the unity of the living world of creation, is commonplace in insular Celtic tradition and serves to invest a given deity or heroic demigod with the attributes traditionally ascribed to certain birds and animals. For instance, the bond between animal and human is implicit in the archetype of the divine swineherds, who are doubtless avatars of the great herdsman god. Further, the Brown Bull of Cuailnge cannot be wholly dissociated from the Tarvos Trigaranus (The Bull of the Three Cranes), pictured on reliefs from Trèves and Notre-Dame-de-Paris and presumably the subject of a lost Gaulish narrative. Among the Celts, as among many other cattle-rearing peoples, the bull was a vivid symbol of power and fertility and appears frequently as a trope in the eulogy of the medieval Irish court poet. It is hardly surprising, therefore, that a god representative of royal and heroic functions should have been represented by this image. Donnotarvos (Brown Bull), the king of the Helvetii mentioned by Caesar, bore a name of great mythic resonance among the Celts, most probably derived from the same deity who appears in the Irish saga as the Brown Bull of Cuailnge.

The animal connections of the Celtic gods are extensive and varied. The iconography shows Cernunnos (The Horned One) associated with the stag, the ram-headed serpent, the bull, and, by implication, with the whole animal world. The iconography also includes boars, horses, dogs, and bears, as well as fish and various kinds of birdsall connected more or less closely with certain deities. This rich diversity is reproduced in even greater abundance in the insular tradition, creating a complex web of connotations and relationships that defy any neat classification. For example, the boar is quite well represented in Celto-Roman sculpture, as in the figure from Euffigneix, Haute-Marne, of a god carrying a boar before him. In insular literature it appears almost ubiquitous. It sometimes leads its pursuers into the otherworld, and often it is in fact a human who was transformed through some mischance or misdeed. Pork was the choice food of the Celts, and, appropriately, in Irish tales the unfailing food of the otherworld is a pig, which, although cooked each night, remains alive and whole each morning.

The horse, index and instrument of the great Indo-European expansion, has always had a special place in the affections of the Celtic peoples. Sometimes in insular tradition, particularly in folk tales, he is the bearer of the dead to the otherworld, a role probably reflected in some monuments in southern Gaul, such as the frieze of horses' heads on a lintel from the Celto-Ligurian sanctuary of Roquepertuse, Bouches-du-Rhone. Epona (from *epos, meaning horse ) was an important Celtic deity and was particularly favored as patron of the cavalry of the Roman army. She has insular analogues in the Welsh Rhiannon and in the Irish Edaín Echraidhe (echraidhe, meaning horse riding ) and Macha, who outran the fastest steeds. There was also a Dea Artio (as well as a Mercurius Artaios), whose name connects her with the bear (Irish, art, meaning bear ); a little bronze group from Bern shows her seated before a large bear with a basket of fruit by her side. Dea Arduinna, who appears seated on a wild boar, may be compared with the Irish goddess Flidhais, who ruled over the beasts of the forest and whose cattle were the wild deer.

Gaulish monuments that show a god or goddess with two or more birds seated on their shoulders call to mind the supernatural birds that are a familiar feature of insular tradition, in which some deities assume bird form occasionally; others, like the war goddesses, do so constantly. The insular catalog of bird imagery is endless. King Conaire's supernatural father came to his mother in bird form, Fann and Lí Ban came to Cú Chulainn as two birds joined by a golden chain, emissaries from the otherworld. Indeed, such wondrous birds are a recognized symbol of the supernatural world. Examples include the three birds of the Irish goddess Cliodhna with their magic song and the three birds of the Welsh Rhiannon who "wake the dead and lull the living to sleep." They all form part of that rich imaginative intuition that envisaged animals, birds, and the whole domain of nature as a mediating element between gods and men and that underlies Celtic literary tradition as well as the fluid discipline of early Irish art.

Invasions of Gods and Men

When Irish monastic scholars began recording native mytho-historical tradition, probably in the second half of the sixth century, they experienced the same difficulty that Christian historiographers have encountered elsewhere in dealing with traditional sources: how to resolve the conflict between Christian and native versions of cosmic origins. Their solution was the familiar one of substituting the biblical doctrine for the earlier part of the native legend, so that it would seem that the legend derived from the doctrine. The fact that the scholars controlled the art of writing invested their new composite historyincrementally elaborated under the influence of the chronicles of Orosius (c. 385420) and Eusebius of Caesarea (c. 260c. 330 ce) and Isidore of Seville's (c. 560636) Etymologiae with an authority it might not otherwise have acquired so quickly. As Christian scholars developed an increasingly close accommodation over the next few centuries with the custodians of native learning, the filidh, their revised version gradually won universal acceptance. Although it did not erase all trace of the earlier tradition, it cancelled out the substance of the original cosmogonic myth. For instance, although the primary ancestral role of Donn, Nuadhu, and others was not forgotten, Adam was accepted as the progenitor of mankind.

The Book of Invasions

The formulation of this revised teaching is attested in poems of the seventh century or earlier, but it was in the twelfth century that it reached its culmination in the pseudohistory entitled Leabhar Gabhála Éireann (The Book of the Taking of Ireland), commonly known as the "Book of Invasions," a cumulative enterprise that carried the tale of Ireland's history from Noah to the Norman conquest. The "taking" in question evidently refers to the coming of the Gaels (or Goidels), but in the extant compilation this is preceded by five other immigrations. The first came before the Flood and was led by either Cesair, a daughter of Bith, who was a son of Noah, or by Banbha, one of the eponyms of Ireland. But the only one to survive the Flood was Fintan (The White Ancient One), who outlived innumerable generations until finally in the Christian period he bore witness to the events of the distant past. The next two settlements were led by Partholón and Nemhedh, respectively. During both, various crafts and social practices were introduced, many lakes were formed, and plains were cleared. These advances indicate in the familiar manner of myths of beginnings how Ireland attained the reality of permanent morphological definition in those times. Both peoples had to withstand the attacks of the Fomhoire, a race of demonic beings who from their haunts beyond the sea posed a perpetual threat to the existence of ordered society.

The main innovations credited to the fourth settlement, comprising the Fir Bholg, the Gailióin, and the Fir Dhomhnann, were sociopolitical in character. By dividing the country into five they instituted the provinces (literally, fifths in Irish), and they introduced the concept of sacred kingship and the relationship between the justice of the king and the fertility of the land. They were followed by the Tuatha Dé Danann (The Tribes/Peoples of the Goddess Danu), who came skilled in the arts of druidry and magic. They brought with them four talismans: the Stone of Fál, which shrieked under the true pretender to kingship; the spear of Lugh, which ensured victory; the sword of Nuadhu, which none escaped; and the caldron of the Daghdha, from which none went unsatisfied. They defeated the Fir Bholg in the First Battle of Magh Tuiredh, but soon they had to take up arms against the Fomhoire.

The Second Battle of Magh Tuiredh

There is also an independent account of the Second Battle of Magh Tuiredh in a text that is perhaps the single most important source for Irish mythology. In it the genesis of the conflict is traced to the First Battle of Magh Tuiredh, in which Nuadhu, king of the Tuatha Dé Danann, lost his arm. Because a personal defect, physical or moral was incompatible with the notion of true kingship, he was obliged to abdicate and was succeeded by Bres (The Beautiful), who had been fathered by Elatha, a king of the Fomhoire, with a woman of the Tuatha Dé, among whom he was reared. But his rule brought only hardship and oppression for the Tuatha Dé, and there was an end to the generosity and hospitality that characterized a true king. Finally he was lampooned by the poet Coirbre in the first satire composed in Irish, and he was asked to give up the kingship. His response was to go to the Fomhoire to seek their support.

Meanwhile, Nuadhu was fitted with a silver arm by Dian Cécht (The Leech) and restored to sovereignty, and from that time forth he was known as Nuadhu Airgedlámh (Nuadhu of the Silver Arm). But when Lugh came to the royal court of Tara and gave proof of his mastery of all the arts, Nuadhu immediately gave way so that Lugh might lead the Tuatha Dé to victory. In the battle itself Lugh called on all the preternatural powers of the craftsmen and magicians of the Tuatha Dé, while Dian Cécht used his own healing magic to revive the slain. The dreaded Balar of the Fomhoire had a "baleful eye" which could destroy armies, but Lugh struck it with his slingstone and killed him. The Fomhoire were then expelled from Ireland forever, and Bres himself was captured, but his life was spared on condition that he divulge to the Tuatha Dé the proper times for plowing, sowing, and reaping.

The Gaels and the Tuatha Dé

The primary subject of the "Book of Invasions" was perhaps the final settlement of prehistoric Ireland, that of the Gaels, or Irish Celts. Because its underlying purpose was to biblicize the origins of the Gaels, it began, as it were, at the beginning, following them in their long journey from Scythia to Egypt and to Spain, whence they finally came to Ireland under the leadership of Míl Espáine (Míl of Spain). The account of this early odyssey is a learned fiction modeled on the story of the wandering of the Israelites in the book of Exodus. But as the narrative approaches Ireland, it undergoes a sea change and begins to draw more overtly on native tradition. The crucial role in the landing is assigned to the poet-seer and judge Amhairghin. By virtue of his wisdom and his mantic power he overcomes the opposition of the Tuatha Dé and becomes the first Gael to set foot on Irish soil. As he does soon the Feast of Beltene (May Day)he sings a song of cosmic affirmation in which he subsumes within himself the various elements of the created universe: "I am an estuary into the sea / I am a wave of the ocean / I am the sound of the sea / I am a salmon in a pool / I am a lake in a plain / I am the strength of art." Like Ka in the Indian tradition and Taliesin in the Welsh, he embodies the potential of all creation, and the timing of his song is particularly appropriate and decisive. Sung as he arrives at the land's edge from the ocean of nonexistence, his words are the prelude to the creation of a new order of which he is the shaper and the source. Through them and through the judgments he pronounces in the succeeding narrative, the Ireland of history is summoned into being.

Having defeated the Tuatha Dé, the Sons of Míl go to the royal center of Tara and on the way meet the three divine eponyms of IrelandBanbha, Fódla, and Ériu. At Tara the three kings of the Tuatha DéMac Cuill, Mac Cécht, and Mac Gréineask for a respite before surrendering sovereignty. Significantly, they refer the conditions to the judgment of Amhairghin. He decides that the Sons of Míl should re-embark and retire beyond the ninth wave, which for the Celts constituted a magic boundary. But when they try to land again, the Tuatha Dé create a magical wind that carries them out to sea. Then Amhairghin invokes directly the land of Ireland, and immediately the wind abates. The Sons of Míl come ashore and defeat the Tuatha Dé at Tailtiu, site of the annual festival instituted by Lugh.

Although defeated, the Tuatha Dé still use their magic powers to extract a reasonable settlement from the Gaels. They agree to divide the country into two parts, the lower half going to the Tuatha Dé and the upper half to the Gaels. Thus is explained the traditional belief that the ancient godsthe sídheóga (fairies)lived underground in sídhe, or fairy mounds. That this belief was traditional already in the seventh century is evidenced by Bishop Tírechán, biographer of Saint Patrick, who noted that the sídh, or gods, dwell in the earth.

Gods of Britain

Early Welsh literary tradition, like the medieval Welsh language, seems further evolved from its archaic roots than its Irish counterpart. This is probably due partly to the cultural effects of the Roman colonization of Britain from the first to the fifth century and partly to the late redaction of the extant material, particularly the prose. But whatever the causes, the result is that Welsh mythological narrative, although preserving some remarkably archaic elements, nonetheless lacks the extensive context found in Irish narrative and betrays the hand of a later redactor or redactors not wholly familiar with the mythological framework from which their materials derived.

Family of Dôn

The main source for Welsh mythological tradition is the collection of tales known as the Mabinogi or Mabinogion, especially the group known as the "Four Branches." These four tales, which were probably redacted toward the end of the eleventh century, take the gods of Britain as their dramatis personae. The last of the four, "Math Son of Mathonwy," deals in particular with the group of gods sometimes referred to as the family of Dôn. The Math of the title is lord of Gwynedd in north Wales. His peculiarity is that he must keep his feet in a virgin's lap except in time of war. When his virginal foot-holder is violated by his sister's sonGilfaethwy, son of Dônwith the connivance of his brother Gwydion, son of Dôn, Math turns the two brothers into male and female animalsstags, boars, and wolvesfor three years, during which time they give birth to three sons.

Subsequently, Math seeks a new foot-holder, and Gwydion suggests his sister, Aranrhod, daughter of Dôn. Math asks her to step over his magic wand as a test of her virginity, and as she does so, she drops a yellow-haired boy and something else, which Gwydion promptly conceals in a chest. The boy is baptized Dylan and immediately makes for the sea and takes on its nature, for which reason he is henceforth called Dylan Eil Don (Dylan son of Wave). The object concealed by Gwydion turns out to be another male child, who in due course is given the name Lleu Llaw Gyffes (Lleu of the Skillful Hand). The rest of the tale is taken up with Lleu's relations with his mother, Aranrhod, and with his beautiful but treacherous wife, Blodeuwedd (Flower-aspect), who had been created for him by Gwydion from the flowers of the oak, the broom, and the meadow sweet. The name Lleu is, of course, the cognate of the Irish Lugh and the Gaulish *Lugus.

The same tale refers incidentally to Gofannon, son of Dôn (Divine Smith), whose name is cognate with the Irish Goibhniu. There is mention elsewhere of Amaethon, son of Dôn, the divine plowman, and there are various references in medieval poetry that indicate the existence of extensive oral tradition about the family of Dôn. Their communal association with magic is reminiscent of the Irish Tuatha Dé Danann, and it has been suggested that Dôn is the equivalent of Irish Donu (Mother of the Gods), the original form of the name Danann.

Family of Llŷr

The three members of the family of LlŷrBranwen, Bendigeidvran (Bran the Blessed), and Manawydanappear in the "Second Branch" of the Mabinogi, although it is only in the "Third Branch" that Manawydan assumes an independent role. The tale is dominated by the enormous figure of Bendigeidvran. When his sister Branwen is ill treated in Ireland, where she has gone as the wife of Matholwch, king of Ireland, he goes with an army to exact vengeance. The British gain victory in a fierce battle with the Irish, but only seven of them survive beside Bendigeidvran, who is wounded in the foot by a poisonous spear. He commands his companions to cut off his head and to bury it at the White Mount in London as a safeguard against invasions. They set out for London and on the way enjoy two periods of otherworldly peace and joy in the presence of his uncorrupted head, at Harlech and on the isle of Gwales.

Clearly, the children of Llŷr are not comparable with those of Dôn: in no sense do they form a pantheon of deities; indeed, Branwen's antiquity is not beyond question. But the association of Bran (as Bendigeidvran was known earlier) and Manawydan is old, and there is an early verse reference to them presiding together over the otherworld and its feast. Manawydan's Irish counterpart is Manannán mac Lir (Son of the Sea), and it is a curious and perhaps significant coincidence that Manannán figures with an Irish Bran in an early lyric tale, which tells of a journey made by Bran to the otherworld. But Manannán is represented as god of the sea, probably replacing the god Nechtan in this role, whereas Manawydan has no such function in Welsh in the extant Welsh texts.

Pwyll, Rhiannon, and Pryderi

In the "First Branch" of the Mabinogi, Pwyll, Lord of Dyfed in southwest Wales, comes to the aid of Arawn, king of Annwn, by slaying his otherworld enemy Hafgan in a single combat that is, in fact, an ordeal by battle of the kind known in early Irish as fír fer (truth of men or heroes). As a result he is henceforth known as Pwyll the Head of Annwn. The Mabinogi represents him here as a mortal, but because his name literally means wisdom and because he is designated Lord of Annwn (the Otherworld), it is probable that he was originally a deity. The latter part of the tale is concerned with the death of the hero Pryderi. Pwyll marries the lady Rhiannon, who first appears to him riding a white horse, and from their union Pryderi is born. But the newborn child is mysteriously abducted, to be discovered later by Teyrnon, Lord of Gwent Is-coed, and reared by him and his wife for several years until they realize the child's true origins and restore him to Pwyll and Rhiannon. After Pwyll's death Pryderi succeeds to the lordship of Dyfed. Later, in the "Third Branch," Rhiannon becomes the wife of Manawydan.

The above merely sketches a complicated narrative whose reference to the underlying mythology is extremely difficult to decipher with any confidence. Teyrnon's name (from *Tigernonos ; Great/Divine Lord) implies a more important role than the one he plays in the tale and, in fact, is a more appropriate title for the lord of the otherworld. Rhiannon (whose name derives from *Rīgantona ; Great/Divine Queen) may be an equivalent of Epona, the Celtic horse goddess, whereas Rhiannon and Pryderi seem to offer a parallel to the pairing of Modron (Great/Divine Mother) and Mabon (Great/Divine Son). The problem is similar to that posed by much of the Welsh mythological evidence in the medieval poetry and the collections of triads: There are numerous references to mythological persons, objects, and events, but these appear without sufficient accompanying matter to set them in context.

Goddesses of the Insular Celts

In The Aran Islands (1907) John M. Synge said of the Aran islanders of the beginning of the twentieth century that they were interested in fertility rather than eroticism, and on the evidence of the extant monuments and literature, his observation could apply to those people who created the mythology of the Celtic goddesses. The Celts had no goddess of love, and so far as one can judge from insular tradition, the numerous sexual liaisons of the goddesses were generally motivated by ritual or social causes, not by erotic ones. Their sexuality was merely the instrument of their fertility, whether in terms of progeny or of the fruitfulness of the land with which they were so often identified.

The cult of the mother goddess, attested in Gaul from prehistoric times, underlies a great deal of Irish and Welsh tradition. The "Second Branch" of the Mabinogi describes Branwen daughter of Llŷr as "one of the three great ancestresses of Britain." The other two presumably are Rhiannon and Aranrhod, and it is clear from Irish literature that the typical goddess figure was often esteemed as the genetrix of peoples. Her personification of the earth tended to be defined and delimited by cultural and political boundaries: The eponymous triad of Ériu, Fódla, and Banbha represent both the reality and the concept of Ireland in its totality, but a multitude of analogous characters also exist that are connected with lesser areasa province, a district, or a particular locale. Some of the latter, such as Áine, Aoibheall, and Cliodhna, have retained their niche in popular tradition and in place names to the present day. In this domain the supernatural female often becomes a dominant figure overshadowing her male counterpart.

One of the most enduring myths of the Celts was that of the solemn union between a ruler and his kingdom, in which the kingdom is conceived in the form of a divine woman. It appears, slightly veiled, in the Arthurian romances and may be reflected at times in the frequent pairing of god and goddess in Celto-Roman sculpture, but its influence is most profound and most widely documented in Irish tradition. The normal way of reporting the inauguration of a king was to say that he was married to (literally, "slept with") his kingdom. From the hundreds if not thousands of references and allusions to this theme, one gains some idea of the ritual union of king and consort as it must have been performed before the effective Christianization of the political establishment in the sixth century. The ritual union had two main elements: first, a libation offered by the bride to her partner, and second, the sex act. The divine nature of Queen Medhbh of Connacht is evidenced by her name as well as by her actions: She who was famed for the number of her successive husbands was called Medhbh (The Intoxicating One), and, under the slightly variant name Medhbh Lethdherg, it was said of her that "she would not permit a king in Tara unless he had her for his wife." The central element was the sexual meeting, and its profound significance is brought out in countless poems and narratives in which the woman is transformed from repulsive age and ugliness to radiant youth and beauty by the act of intercourse with her ordained mate.

As leader of the Connacht armies, Medhbh is associated with war as well as with sovereignty, but, in general, the warlike aspect of the goddess is manifested indirectly: she influences the fortunes of war rather than actually participating. Other goddesses teach the art of fighting, including Buanann (The Lasting One); Scáthach (The Shadowy One), from whom Cú Chulainn acquired his heroic skills; and the formidable trio of Morríghan (Phantom Queen), Bodhbh (Scald-Crow), and Nemhain (Frenzy) or Macha, who haunt the battlefield to incite the fighters or to hinder them by their magic. These had their equivalents throughout the Celtic world: The name Bodhbh Chatha (Crow/Raven of Battle) is the exact cognate of Cathubodua, attested in Haute-Savoie, and the trio of war goddesses recurs in Britain at Benwell in the inscription "Lamiis tribus " (to the three Lamiae).

In direct contrast to these ruthless furies are those charming women who inhabit the happy otherworld in such numbers that it came to be called Tír inna mBan (The Land of Women) in some contexts. Sometimes they come as emissaries from the land of primeval innocence where the pleasures of love are untainted by guilt and where sickness and disease are unknown. Conla son of Conn is induced to go there by "a young and beautiful woman of noble race whom neither death awaits nor old age," and Bran son of Febhal is similarly persuaded by a woman bearing a silvery branch from the wondrous apple tree, which is a characteristic feature of the Celtic otherworld. The multiforms of the insular Celtic goddesses are endless, and sometimes the named figure changes her role from one context to another. For example, in Mythe et épopée (1968), Georges Dumézil has sought to demonstrate from three separate tales that the goddess Macha, eponym of the old pagan center of Emhain Mhacha and of the Christian metropolis of Ard Macha (modern Armagh), reflects in her several roles the Indo-European trifunctional system of religion, warrior prowess, and fertility. Although his argument is open to question, it is nonetheless true that several of the prominent goddesses have widely varying epiphanies.

Mythic Space and Time

In a tradition in which the natural and the supernatural realms frequently converge, it is not surprising that there is a constant awareness of the relativities of time and space. This is particularly true of texts relating explicitly to the otherworld, but it is common throughout Irish and much of Welsh literature. The land of Ireland itself, with its place names and physical features, seems to shift with enigmatic ease between the two levels of perception. The early redactors of the written texts were fascinated by the contrasting effects of changing perspective, as when the god Manannán describes the sea as a flowery plain or the monks of Clonmacnois observe a boat sail in the sky over their head and drop its anchor by their church door.

But certain places are permanently set apart from their secular environment: cult sites, the precincts of sacred festivals, and, above all, the notional center of the ethnic world of native tradition. This concept of the center is one of the constants of Celtic ideology, and it retained a good deal of its ancient symbolism in Irish learned literature as late as the seventeenth century. Caesar reports that the Gaulish druids assembled each year at a holy place in the lands of the Carnutes, which was regarded as the center of Gaul. His term locus consecratus may well translate the word nemeton (sacred place),which is found in place-names throughout the Celtic world. According to Strabo (c. 63 bce24 ce), the Council of the Galatians met at a place known as Drunemeton (Oak Sanctuary). In Ireland the druids were closely associated with Uisnech, the "navel" of Ireland, the location of the primal fire, and reputedly the site of a great festival. The focus of sacral kingship was at Tara in the central province of Midhe (Middle) and it was entirely fitting that St Patrick's late seventh-century biographer, Muirchú maccu Machtheni, who describes Tara as caput Scotorum (the capital of Ireland), should have him travel there to demonstrate his superiority over the druids of Loegaire mac Néill, imperator barbarorum and "ancestor of the royal stock of almost the whole of this island."

The great social assemblies of ancient Ireland were generally held at one of the seasonal festivals. The Irish year, like the Indo-European year, was divided into two halves, samh (summer) and gamh (winter). The summer half began at Beltene or Cédshamhain, the first of May, and the winter half at Samhain, the first of November. These halves were further subdivided by the quarter days of Imbolg, the first of February and the beginning of spring, and Lughnasadh, the first of August and the start of the harvest festival associated with the god Lugh. The old binary division is found also in the famous bronze calendar discovered at Coligny, near Bourg, which probably dates from the early first century ce or late first century bce. Judging from the calendar, the Gaulish druids divided the year into two halves beginning with the months Samon(i-) and Giamon(i-). Of the two names for the beginning of summer, Beltene may have referred originally to the fire ritual traditionally held at that time: bel- probably means shining or bright, and tene may be related to the Irish word for fire. In the course of time, however, Beltene displaced the older term Cédshamhain or Cédamhuin (cf. the Welsh cognate Cyntefin) as the name for the festival season itself.


In Caesar's time the institution of kingship was already on the way to dissolution in Gaul, having been widely displaced by the secular office of vergobret (chief magistrate), although it is clear from the extant evidence that all the tribal territories, the civitates of Caesar's time, had earlier been ruled by kings in the mold of those of early Britain and Ireland. The medieval Irish king tales inevitably share in some degree the values of the general heroic literature, but these are not their main preoccupation. They are concerned rather with the affirmation of political and social realities and with the safeguarding of traditional institutions: the status and functions of the king and the sacred ritual of inauguration that set the seal on his accession to power, the origins of tribes and dynasties and exemplary tales of their internecine conflicts, the deeds and judgements of famous rulers of the past, and so on. The sacral kingship was both the pivot and the foundation of the social order, and the king was its personification. If his conduct or even his person were blemished in any way, the effect of his blemish would be visited on his kingdom, diminishing its integrity and prosperity; conversely, fortune favored the righteous ruler and his people flourished and his territory became rich and fertile. As the instrument of justice, the king must be seen to be fair and flawless in his decisions and several of the famous kings of legend are frequently presented as models of regal wisdom and justice. Thus, Cormac mac Airt is pictured as a paragon of kingship and as an Irish Solomon. His accession came about when he proposed a just judgement after his predecessor Lughaidh mac Con had been deposed for delivering an unjust one. Conaire Mór is likewise an exemplary king whose reign brings peace and well-being to the land until he tempers justice with excessive mercy in the case of his three marauding foster brothers. Immediately a train of events is set in motion that leads inexorably to his death in a welter of violence.

As the central pillar of his kingdom the sacral king was its primary point of contact with the world of the supernatural in pre-Christian time, and as such it was necessary to insulate him from harmful intervention from whatever source. Thus each of the five provincial kings was subject to a set of gessa (taboos), which made manifest the transcendent nature of his role and were presumably intended to hedge him from unnecessary danger. When, however, as in the case of Conaire Mór, he unavoidably or unwittingly violates his gessa, he is already doomed to disaster and death. The crucial touchstone of a king's reign was the fír flathemon (the ruler's truth/righteousness) with which he discharged the responsibilities of his office. The analogy between the fír flathemon and the Indic "act of truth" has long been recognized and there is acceptance that together they represent an Indo-European institution. The concept of the Ruler's Truth is referred to frequently in Irish literature, most notably in Audacht Morainn (The Testament of Morann, a legendary law-giver), an early example of the literary genre of the speculum principum (literary, "mirror of princes"), which was designed to give counsel and guidance to a king. The Audacht was probably written toward the end of the seventh century ce, but the genre was already long established in oral tradition, and it is widely accepted that the European speculum principum derives partly from the Irish model and that the Audacht itself contains much that is referable to Irish kingship in the pre-Christian period.

As a genre the speculum was evidently associated with the rite of royal inauguration and was probably recited publicly by a druid or fili in the course of the ritual ceremony. In the pre-Christian and early Christian period, as reflected in the classical law tracts, there were three grades of kingship: the rí tuaithe (king of a tuath; literally, "people " or "tribe "), the smallest political entity; the ruiri (great king or overking), who, as well as ruling over his own petty kingdom, received tribute from several other tuatha; and finally the rí ruirech (king of overkings), who is equated to the rí cóicid (king of a province). Despite the wide disparity of these kingships in range and importance, each of them had its own sacred king and its own inauguration site. However, it is clear that Taraas the ideological focus of sacral kingship and at the heart of the Irish cosmographic systemenjoyed a special prestige as a kind of primus inter pares (first among equals) among royal sites and thus became the goal, real or notional, of ambitious kings throughout the early Middle Ages. Feis Temhra (The Feast/Wedding Feast of Tara) was the great festival held in pagan times to confirm a new king and to celebrate his ritual marriage to his kingdom. At Tara stood the Lia Fáil (Stone of Fál), the "stone penis" that cried out when it came in contact with the man destined to be king. Feis, verbal noun of the verb foaid, means literally "to sleep, spend the night," and, in the context of the royal confirmation, it refers to the ritual marriage of the king and his kingdom, as underlined in the alternative expression banais rígi (wedding feast of kingship), in which banais is compounded of ben (woman) and feis. This terminology continues to be used of various royal inaugurations in annalistic and other texts, even in the Anglo-Norman period. One can only speculate as to the precise form the marriage ritual may have assumed in pre-Christian timesactual union with a surrogate bride or a simulated union that included the proffering of the drink of sovereignty. The earliest list of reigning kings for the kingship of Tara is furnished by the seventh-century text Baile Chuind (The vision of [King] Conn [Cétchathach]), which purports to prophesy the individual kings who were to reign in Tara from the time of his son Art onward. Its literal formula for "X shall reign" is "X shall drink it," in which the formal potion presented to the ordinand is employed as a synonym for the combined ceremony of sacral investiture and the exercise of kingship. The text is devoid of explanatory introduction and is presumably to be understood as spoken by Conn himself, but when it was reworked and expanded in a more narrative and iconically stylized context in the ninth century in the tale Baile in Scáil (The phantom's vision), the prophecy is spoken by the god Lugh, the Irish (and Celtic) divinity traditionally regarded as personifying the ideal of kingship. It tells how Conn went on a circuit of the rampart of Tara accompanied by his three druids to guard against hostile incursions by forces from the otherworld, perhaps a reference to the familiar taboo that forbade the king to let the sun rise on him in Tara. One recalls, for example, the story of Aillén mac Midgna from the otherworld mound of Síd Finnachaid who came regularly to Tara at Samain (Hallowe'en), lulled its people to sleep with his supernatural music (ceol sídhi ) and burned it down with a pillar of fire, until finally he was slain by the leader of the Fiana, Fionn mac Cumaill. So, when Conn mounts the rampart of Tara in Baile in Scáil, he comes into direct contact with the otherworld, although, in this instance, under one of its more benign aspects. A magic mist enveloped the king and his companions and a horseman (the scál or phantom) approached and asked them to accompany him to his dwelling. Within they found a girl seated on a chair of crystal and wearing a golden crown. Beside her stood a vessel of gold with a golden cup nearby. The phantom, seated on his throne, identified himself as the god Lugh and declared that he had come to announce to them the names of Conn's successors and the duration of their reigns. The young woman was the sovereignty of Ireland and when she asked to whom she should offer the cup of red ale (dergfhlaith ), the phantom enumerated his catalog of the kings who would follow Conn.

The terminology used in reference to the hieros gamos (sacred marriage) of king and goddess points to some sort of sexual union taking place in pre-Christian times, as do the several tales of the loathsome hag who is transformed to youth and beauty by intercourse with the rightful candidate for kingship, a theme that is exploited for political dynastic ends in extant medieval versions. But accounts of the actual inauguration ceremony are of later date and betray some degree of ecclesiastical influence. Inevitably the Christian Church, conscious of the pivotal significance of the sacral kingship to native society, sought to arrogate to itself a central role in "ordaining" the ruler, and thus to sanitize the most incompatible elements of the traditional ritual. But tradition was tenacious. According to a quite late prose account (fourteen to sixteenth century) of the ceremonial inauguration of the Ó Conchubhar kings of Connacht, many clerics and all the subkings of the province were present, yet it was Ó Maoil Chonaire, the fili, modern proxy of the ancient druid, who installed him as king (aga ríghadh ) by presenting him with the rod of sovereignty, and, the text adds, none but Ó Maoil Chonaire had the right to be with the king on the inauguration mound apart from the keeper of the mound.

Moreover, the gradual revision of the inauguration ceremony during the pre-Norman centuries may not have proceeded as regularly and universally as most later accounts might suggest. In a well-known passage of his Topographia Hiberniae, which is based on information garnered during his stay in Ireland in the late twelfth century, Gerald of Wales describes a "barbarous and abominable" rite of inauguration practiced in what is now County Donegal. A white mare is brought to the midst of the assembled people, the future chief has sexual union with the mare, which is then killed, cut in pieces, and boiled. The chief then sits in this bath, eats of the mare's meat and drinks of the broth, and thus kingship and power is conferred on him. Despite the lack of supporting native testimonies, it is difficult to discount the striking analogy this bizarre ritual presents to the Indic asvamedha (horse sacrifice), one that is accepted by most comparatists. The main disparity is that in the Irish version the sex act involves the king and a mare instead of the queen and a stallion, as in India, but some scholars would, in fact, argue that the Indo-European inauguration was primarily between king and mare. However, even if the essential authenticity of Gerald's account is accepted, it does not follow, as some have assumed, that such a rite was practiced in or close to his time. Elsewhere he draws on reportssome fabulous, others more factualgathered from a variety of sources, oral as well as written. In this particular instance, it is a piece of seanchas (oral history), referring to an already more or less obsolete era. Nonetheless, it is a useful reminder that the version of native belief and ideology mediated to modern readers by the redactors of the medieval monasteries is less than comprehensive.

Another archaic institution associated with royal inauguration was the crech ríg (royal foray), which is still attested in the post-twelfth century Anglo-Norman period. As in ancient India, such a cross-border raid was a recognized occasion for the new king to demonstrate his fitness for office and at the same time to acquire the means to make appropriate show of his largesse.

The heroic ideal: The Ulster Cycle

Like the sacral king of prehistoric tradition, the hero occupied an ambiguous status between god and men. Typically, he has a divine as well as a human father, and his trials and achievements bring him into contact with supernatural powers more frequently than other mortals. He has many incarnations in insular Celtic literature, but it is above all the Ulster Cycle that represents him in the quintessential heroic setting.

The cycle is set in the province of Ulster when it was dominated by the Ulaidh, the people from whom the province derived its name, at a time somewhere between the coming of the Celts, perhaps as late as the third century bce, and the conquest of the Ulaidh, which may have taken place in the early fifth century ce. The cycle portrays an aristocratic warrior society with a La Tène (Second Iron Age) type material culture, and in many respects the society shows striking correspondences with what is reported of independent Gaul. The king of the Ulaidh at this time was Conchobhar mac Nessa, who had his royal court at Emhain Mhacha near the present city of Armagh. He presided over a numerous company, which included the youthful Cú Chulainn, the senior heroes Conall Cernach and Ferghus mac Roich, and such others as the druid Cathbhadh, the wise peacemaker Sencha mac Ailella, and the inveterate mischief-maker Bricriu, known as Nemhthenga (Poison-tongue). These characters constitute the cast of an extensive literature of which the centrepiece is the great saga Táin Bó Cuailnge (The cattle raid of Cuailnge). It tells of Queen Medhbh of Connacht's incursion into Ulster with the object of seizing the great Brown Bull of Cuailnge, which was of divine origin. As a result of a curse by the goddess Macha, the Ulstermen are unable to resist the attack, and it falls to the young Cú Chulainn to defend the province single-handedly. By engaging in a series of single combats with heroes of the Connacht army, he hinders their advance until the Ulstermen recover their strength and rout their enemies. The climax and finale of the tale is the tremendous encounter in which the bull of Cuailnge slays the Finnbhennach, the white-horned bull of Connacht.

As the heroic milieu par excellence, the court of Conchobhar at Emhain became the focus for a wide variety of tales reflecting the different facets of the heroic ethos, and as the quintessential hero Cú Chulainn became the subject of many narratives exploring the nature of the hero's mediating role between gods and men and his singular relationship with his own community. Cú Chulainn experiences the perennial dilemma of the supreme hero caught in the insoluble contradictions of his ambiguous status. Neither divine nor merely human, Cú Chulainn lives within the tribe and yet does not wholly belong; a member of a heroic confraternity, he characteristically stands alone. His initiation to the heroic circle is recounted in a section of Táin Bó Cuailnge that narrates his boyhood deeds (macghnímhartha ), which, linguistically, is not part of the oldest stratum of the text (it may belong to the ninth century), although its content is part of an archaic tradition. For his first exploit, Cú Chulainn slays the three fearsome sons of Nechta Scéne who have been a scourge on the Ulstermen. Here the narrative appears to reproduce an old lndo-European motif of the hero's victory over a trio of adversaries or a three-headed monster. He also, for the first time, experiences the riastradh (grotesque distortion) and the phenomenal body heat that are the external manifestations of his battle fury and that mark him in Irish tradition as a hero above heroes. These traits also have old and widespread analogues.

Cú Chulainn's career is a short one, but because it constitutes a paradigm of the hero, the mythmakers and storytellers have taken the critical stages of his life and woven a web of narrative around each: his threefold birth distinguished by incest and divine paternity, familiar marks of the sacred conception of the hero; his martial training with the otherworldly Scáthach; his wooing of Emher and his marriage; and finally his death, which, because he was invincible by merely human means, could only be effected through trickery and sorcery. This framework has also accommodated a number of other more occasional tales, such as those of his adventures in the otherworld or the tragic Aided Aenfhir Aífe (The death of Aífe's only son), which brings Cú Chulainn to slay his own son through a combination of moral compulsion and mistaken identity.

But Cú Chulainn and his life cycle are only a part of the larger cycle of the Ulster tales and in many he plays a relatively small role or none at all. His singular importance is that he epitomizes the heroic virtues and values. By the seventh century ce he had become a focus in the written literature for archaic traditions pertaining to what Dumézil defined as the second of the Indo-European social functionsthat of the warrior.

The Fionn Cycle

In early Irish, the Fionn Cycle was also known as the Fianaighecht. It comprises a complex of stories and traditions about the Fian, the band of hunter-warriors led by Fionn mac Cumhaill. The cycle is commonly called the Fenian Cycle, a modern Anglicization, or the Ossianic Cycle, after Fionn's son Oisin (or Ossian). Etymologically, the term fian (plural, fiana ) embodies the notion of living by the hunt or by force of arms, and this notion corresponds exactly with the role of the Fiana in Irish tradition. Originally there were several groups of Fiana, but the fame of Fionn's company relegated the others to obscurity. Each féinnidh (individual member of the Fian) was required to undergo initiatory trials of his skill and endurance before admittance, and once accepted he had to sever his legal and social connections with his kin and his tribe and abandon the associated rights and responsibilities. Yet although he placed himself outside the tribal community, he did not place himself outside the law, for the Fiana were recognized by law and tradition as fulfilling a legitimate function. Many legends picture the Fiana as the defenders of Ireland against the incursions of foreignthat is, in effect, supernaturalenemies. From the eleventh or twelfth century onward, and perhaps even earlier, these enemies are often identified in an ambiguous, mythopoeic (relating to mythmaking) fashion with the Viking raiders of the ninth century.

Some have recognized the Celtic form *vindos (white, fair)the source of Irish Fionn and Welsh Gwynnin the Celtic deity name/epithet Vindonnus, and thus concluded that Fionn himself was originally divine, although this is questionable. *Vindos is related to the Indo-European stem *ui-n-d (finds out, knows). It also has been suggested that Fionn's name means "he who finds out, he who knows." This accords with his role in tradition, which represents him as poet and seer as well as warrior-hunter, perhaps like his Welsh counterpart Gwynn ap Nudd, who appears fleetingly in Welsh tradition as a "magic warrior-huntsman." Fionn is sometimes said to have acquired his supernatural knowledge by tasting the otherworldly liquor. His normal means of divination was simply to chew his thumb, with which he had once touched the Salmon of Knowledge, which he was cooking for his master in poetry and magic. Moreover, poetry and preternatural vision have always been characteristic attributes of the Fionn cycle as a whole.

Like Cú Chulainn, Fionn is also the subject of a narrative recounting his boyhood deeds. His birth followed soon after his father's death at the hands of the rival band of the Sons of Morna. He was reared secretly in the forest by two female warriors until he was ready to assert his precocious claim to the leadership of the Fian. He killed a malevolent being called Aillén mac Midgna, who came each year to burn down the royal court of Tara (one of several variants of a myth in which Fionn figured as conqueror of a supernatural one-eyed arsonist). Even within the Fian his archrival was Goll (one-eyed) mac Morna, also known as Aodh (Fire). There is an obvious analogy here with the myth of Lugh's defeat of Balar, and it has, in fact, been argued that Fionn was simply another name and persona for that deity. However, although Lugh is represented as being closely associated with the sacred function of kingship, Fionn's relationship to kingship is, at the very least, ambiguous. It is true that he and his followers became closely associated with the king of Tara as a kind of standing army, but it has been suggested that this is a fairly late development. Earlier their role as mercenaries appears to have been more marginal and ambivalent.

This marginal status may partly explain why the Fianaighecht was accorded little space in the written texts before the eleventh and twelfth centuries, although it is attested as early as the Ulster Cycle. By and large the literature of prestige such as the Ulster Cycle reinforced the structures and usages of organized aristocratic society within its clearly defined political boundaries. But the Fian's environment was outside and beyond this cultivated domain in the forest and the wilderness. Here they roamed at will, on foot or on horseback, unlike the Ulster heroes, who traveled in chariots. Intimately connected with nature, both animate and inanimate, their world blurred and often dissolved the boundaries of social and natural categories. For example, several of the feinnidi were born of mothers in animal form, and the Fian's great hounds, Bran and Sgeolang, had a human mother. It is hardly surprising that Fian mythology has always had a firm hold on the popular imagination and that it only gained prominence in the written tradition when the learned class began to react to the pressure of sociopolitical change in the eleventh and twelfth centuries.

The ambiguous nature of the region inhabited by the Fian emerges clearly in their relations with the otherworld. Whereas in the Ulster tales the association of the two worlds tends to happen at specific timesat the great calendar festivals, for instance, or during initiation ritualsamong the Fian these associations are casual and continual. The Fian's liminal status ensures that they can participate freely in both the natural and the supernatural world as they are able to easily cross the threshold between worlds. In this as in much else they correspond to the heroes of Arthur's court and there can be little doubt that the cycles of Fionn and Arthur, whatever their later vicissitudes, derive from the same sector of insular mythology.

The "Elopement of Diarmaid and Gráinne," one of the most popular tales in the Fianaighecht, tells how the mature Fionn loses the beautiful Gráinne to Diarmaid ua Duibhne (The Master and Charmer of Women), just as Arthur loses Gwenhwyfar (Guinevere) to Medrawd (Melwas). The tale is one of several Irish analogues of the romance of Tristan and Iseult, and it also ends in tragedy, when Diarmaid is killed by the magic boar of Beann Ghulban with Fionn's connivance. It has been suggested that Gráinne's name, which can mean literally ugliness, obliquely identifies her with the version of the sovereignty goddess who appears as a repulsive hag until she is transformed to youthful beauty by union with her rightful and royal mate. Diarmaid Donn (Brown, Dark) may originally have been the god Donn who ruled the otherworld of the dead.

The most comprehensive source for the Fianaighecht is a long frame story entitled Agallamh na Senórach (The converse of the old men), which was probably compiled near the end of the twelfth century. The title indicates the convenient device on which the massive narrative rests: Caoilte mac Rónáin, one of the principal members of the Fian, long outlives his contemporaries and eventually meets with St. Patrick, who is on his mission of Christianization. Caoilte accompanies Patrick on his journey throughout the Irish countryside and, at the saint's request, tells him the stories associated with its hills, rivers, plains, and other natural features. The result is a vast thesaurus of place-name lore (dinnshenchas ), which brings together the several streams of learned and popular tradition that went into the making of the Fionn Cycle.

System or Chaos

Matthew Arnold admired the Celts for their lyric gifts, but he claimed, perhaps not without some reason, that they lacked the sense of architecture in their literary compositions. It is a sentiment that has been echoed by many students of Celtic religion and mythology when confronted with the frustratingly formless and unfinished character of the rich corpus of evidence. This feeling has been aptly expressed by Marie-Louise Sjoestedt in her Gods and Heroes of the Celts (1949):

In travelling through the dense forest of the insular legends, and stirring the ashes of the continental Celtic world, we did not hope to uncover the plan of a vast edifice, a temple of the Celtic gods, partly overrun by the luxuriant wilderness and partly ruined by invaders. The indications are that this edifice never existed. Other people raised temples to their gods, and their very mythologies are temples whose architecture reproduces the symmetry of a cosmic or social orderan order both cosmic and social. It is in the wild solitude of the nemeton and sacred woodland, that the Celtic tribe meets its gods, and its mythical world is a sacred forest, pathless and unbounded, which is inhabited by mysterious powers. We seek for a cosmos and find chaos. The investigation of the insular tradition leaves one with a sense of something missing. One searches in vain for traces of those vast conceptions of the origin and final destiny of the world which dominate other Indo-European mythologies. Was there a Celtic cosmogony or eschatology? Must we suppose from the few allusions, vague and banal as they are, which Caesar or Pomponius Mela have made to the teaching of the druids, that a whole aspect, and an essential aspect, of this mythical world is hidden from us and will remain hidden? Should we explain the silence of our texts by the censorship of Christian monks, who were nevertheless liberal enough to allow the preservation of episodes much stained with paganism, and features most shocking to the Christian mentality? (p. 92)

Sjoestedt's own reply to this last rhetorical question would have been a clear negative, but some more recent studies suggest a qualified affirmative. In fact, there are grounds for believing that the early monastic redactors, for all their undoubted empathy and tolerance, did censor pagan learned tradition by omission as well as by critical editing, and that their omission most seriously affected those areas in which conflict of doctrines was least acceptable to Christian orthodoxy: ritual, cosmogony, and eschatology.

In 1918 Joseph Vendryes demonstrated in an important article that the Celtic languages, and particularly early Irish, preserve the remnants of an old Indo-European religious vocabulary originating with the hieratic ancestors of brahmans, pontifs, and druids. Since then it has become increasingly clear that these particularities of terminology are not to be seen as isolated fossils but rather as reflecting interrelated elements of a system of socioreligious thought and practice, which must have persisted substantially unchanged until a relatively late date, perhapsin Ireland at leastuntil the establishment of Christianity. The numerous survivals of archaisms from Indo-European ideology, ritual, and liturgy in early Irish recorded tradition strongly support this conclusion. So also does the "deep structure" of early Irish narrative that is gradually being uncovered by the close analysis of individual texts. In the context of such fundamental and constantly recurring themes as the sacral kingship, the king as mediator between the secular and the supernatural world, the antinomy of ideological unity and political fragmentation, and the concept of social or cosmic order, these early texts often reveal a complex weave of structured allusion that presupposes in the not too distant past a coherent and authoritative system of politico-religious and juridico-religious belief and speculation.

However, it would be wrong to assume that the texts offer a complete and consistent record of that system, not merely because monastic redactors practiced conscious censorship and selectivity but also because the texts were recorded long after druidic paganism had ceased to be the official and uncontested religion of the country. By reason of this remove in time and motivation, the early Irish documentation belongs largely to the category to which Georges Dumézil has applied the term mythologie littérarisée It is the concern of contemporary scholars to analyze and interpret this rich documentation and to restate it in mythico-religious rather than literary terms.

See Also

Druids; Fomhoire; Mabinogion; Matres; Sídh; Táin Bó Cuailnge; Tuatha Dé Danann.


Arnold, Matthew. On the Study of Celtic Literature. London, 1867.

Bieler, Ludwig, ed. The Patrician Texts in the Book of Armagh. Dublin, 1979.

Binchy, Daniel A. Celtic and Anglo-Saxon Kingship. Oxford, 1970.

Birkhan, Helmut. Kelten: Versuch einer Gesamtdarstellung ihrer Kultur. Vienna, 1997. A comprehensive and detailed account of ancient and medieval Celtic culture with generous treatment of religion, mythology, and institutions.

Bromwich, Rachel. Trioedd Ynys Prydein: The Welsh Triads. 2d ed. Cardiff, 1978. This edition of the medieval triads and the rich commentary and notes that accompany it are an invaluable source of information on early Welsh and British history, myth, and legend.

Duval, Paul-Marie. Les dieux de la Gaule. Rev. ed. Paris, 1976. A convenient compendium of what is known and surmised about the Gaulish gods.

Dumézil, Georges. Horace et les Curiaces. Paris, 1942.

Dumézil, Georges. Naissance de Rome. Paris, 1944.

Dumézil, Georges. Dieux des Indo-Européens, Paris, 1952.

Dumézil, Georges. Mythe et épopée. Vol. 1. Paris, 1968.

Gray, Elizabeth A. Cath Maige Tuired: The Second Battle of Mag Tuired. London, 1982. An edition of this important mythological text. Gray's "Cath Maige Tuired: Myth and Structure," Éigse 18 (1981): 183209 and 19 (19821983): 135, 230262, presents a detailed interpretative analysis of the content of the tale.

Lambrechts, Pierre. Contributions a l'étude des divinités celtiques. Bruges, 1942.

Le Roux, Françoise, and Christian J. Guyonvarc'h. La Civilisation celtique Rennes: La société celtique: dans l'idéologie trifonctionnelle et la tradition religieuse indo-européennes. Rennes, France, 1991.

Le Roux, Françoise, and Christian J. Guyonvarc'h. Les fêtes celtiques. Rennes, France, 1995.

Lucas, A. T. Cattle in Ancient Ireland. Kilkenny, Ireland, 1989.

Mac Cana, Proinsias. Celtic Mythology. Rev. ed. Feltham, U.K., 1983. A short survey of the subject with illustrations of sculpture, metalwork, and so on.

MacCulloch, J. A. The Religion of the Ancient Celts. Edinburgh, 1911. Reprinted as Celtic Mythology (Boston, 1918). Still useful if read in conjunction with more recent accounts.

MacNeill, Máire. The Festival of Lughnasa. Oxford, 1962. A comprehensive inventory of all the local festivals in Ireland that can be shown to continue the Celtic feast of Lugh, together with a very helpful commentary and a rich collection of texts, largely from the oral tradition.

Maier, Bernhard. "Is Lug to be identified with Mercury? (Bell. Gall. VI, 17,1): New Suggestions to an Old Problem." Ériu 47 (1996): 12735.

Maier, Bernhard. Dictionary of Celtic Religion and Culture. Woodbridge, U.K., 1997. Translation of Lexikon der keltischen Religion und Kultur (Stuttgart, Germany, 1994; reprint, 1997). A useful and accurate work of reference covering the continental and insular evidenceliterature, iconography, archaeology.

Meyer, Kuno, ed. and trans., and Alfred Nutt. The Voyage of Bran, Son of Febal, to the Land of the Living. 2 vols. London, 18951897. Includes a long commentary on the Celtic concept of the otherworld and the doctrine of rebirth. Largely superseded by more recent studies, it still contains many useful insights.

Murphy, Gerard, ed. and trans. Duanaire Finn : The Book of the Lays of Fionn. Vol. 3. Dublin, 1953. Includes a long and valuable commentary on the history of the Fionn Cycle and on the relationship between medieval manuscript and modern oral versions.

Nagy, Joseph Falaky. The Wisdom of the Outlaw: The Boyhood Deeds of Finn in Gaelic Narrative Tradition. Berkeley, Calif., 1983. An excellent interpretative commentary on the Irish Fionn Cycle, the first extended study of the cycle in terms of modern mythological theory. It explores the internal consistency of the cycle as reflected in some of its constituent narratives and brings out the markedly liminal character of Fionn and his followers.

Ó Cathasaigh, Tomas. The Heroic Biography of Cormac mac Airt. Dublin, 1977. A perceptive exposition of the status and function of the Irish hero-king as reflected in the legends of Cormac mac Airt.

O'Flaherty, Wendy Doniger. Women, Androgynes, and Other Mythical Beasts. Chicago, 1980. In particular, see chapter 6, "The Indo-European Mare."

O'Rahilly, Thomas F. Early Irish History and Mythology. Dublin, 1946. Valuable for its coverage of Irish literary resources in all periods and for its brilliant analyses of medieval texts, but sometimes rather outmoded and idiosyncratic in its treatment of essentially mythological narratives as reflections of historical events.

Ó Riain, P. "Traces of Lug in Early Irish Hagiographical Tradition." Zeitschrift für celtische Philologie 36 (1978): 13855.

Ó Riain, Pádraig. "The 'Crech Ríg' or 'Regal Prey.'" Éigse 15 (1973): 2430.

Puhvel, Jaan. "Aspects of equine functionality." In Myth and Law among the Indo-Europeans, edited by Jaan Puhvel, pp. 16969. Berkeley, Calif., 1970.

Rees, Alwyn, and Brinley Rees. Celtic Heritage. London, 1961. An important and stimulating work that seeks to structure insular Celtic tradition in terms of a number of ideological concepts and motivations. It is inspired by the Dumézilian system of analysis, applied in a flexible and imaginative fashion.

Ross, Anne. Pagan Celtic Britain. London, 1967. Surveys the British repertory of images for the Celtic gods and their attributes. Contains an extensive discussion of the several main categories of deity: horned god, warrior god, divine animals, among others. Useful also for its rich comparative documentation from insular literary and folklore sources.

Scowcroft, R. Mark. "Leabhar Gabhála. Part II: The Growth of the Tradition." Ériu 39 (1988) 166. Offers an excellent analytic commentary on the new synthetic mythology that emerged from the fusion of pagan myth and legend with the Latin-mediated learning of clerics and schoolmen.

Sjoestedt, Marie-Louise. Dieux et héros des Celtes. Paris, 1940. Translated by Myles Dillon as Gods and Heroes of the Celts (London, 1949). A short but perceptive survey of Celtic, mainly Irish, mythology and hero tales. At the time of its publication it offered fresh insights into the nature of Celtic myth and is still necessary reading.

Synge, John M. The Aran Islands. Drawings by Jack B. Yeats. Dublin, 1907.

Vendryes, Joseph. "Les correspondances de vocabulaire entre l'indo-iranien et l'italo-celtique." Mémoires de la Société de Linguistique de Paris 20 (1918): 265285.

Vendryes, Joseph. Les religions des Celtes (1948). Revised by Pierre-Yves Lambert. Vol. 1. Spézet, France, 1997. It is primarily an exhaustive catalogue of the varied data, both continental and insular, relating to Celtic religion. More descriptive than theoretical, it is still a useful source of information.

Vries, Jan de. Keltische Religion. Stuttgart, Germany, 1961. A comprehensive treatment of the whole of Celtic religion. It is well documented and strong on Indo-European and other comparative aspects, less so on the insular tradition, although the latter is given fairly generous coverage.

Proinsias Mac Cana (1987 and 2005)